An asset fault locating method based on a knowledge graph

Abstract

The application discloses a kind of asset fault location methods based on knowledge graph, specifically relates to asset fault location technical field, for solving the problem of multiple source alarm cross node time mispositioning causing aggregation distortion;By acquiring asset topology knowledge graph and multi-source time sequence alarm event stream, according to source sampling time and graph processing arrival time, node state fragment and node state time sequence are constructed, state transition record is matched along node connection relationship, state transition matching sequence and edge time sequence compensation relationship are formed;According to the boundary of node state fragment and mapping time advancing relationship, the current available mapping interval is formed, the node state fragment is extracted to form the neighborhood message set, the state fragment extraction order is marked after moving for the node connection relationship that does not meet the extraction condition, and the target aggregation time reference is updated according to the cumulative edge time sequence compensation relationship, the neighborhood message set is supplemented, and the fault representation and root cause node positioning result are obtained.

Images